Effects of TiO2 in Low Temperature Propylene Epoxidation Using Gold

Dec 21, 2017 - ... K. Combined characterizations of the Au-based catalysts by X-ray absorption spectroscopy, scanning transmission electron microscopy...
0 downloads 4 Views 1MB Size
Subscriber access provided by ECU Libraries

Article 2

Effects of TiO in Low Temperature Propylene Epoxidation Using Gold Catalysts Zheng Lu, Mar Piernavieja-Hermida, C. Heath Turner, Zili Wu, and Yu Lei J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b10902 • Publication Date (Web): 21 Dec 2017 Downloaded from http://pubs.acs.org on December 22, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Effects of TiO2 in Low Temperature Propylene Epoxidation Using Gold Catalysts Zheng Lu†, Mar Piernavieja-Hermida†, C. Heath Turner‡, Zili Wu¶ and Yu Lei†,* †

Department of Chemical and Materials Engineering, University of Alabama in Huntsville, AL

35899, USA ‡

Department of Chemical and Biological Engineering, University of Alabama, Tuscaloosa, AL

35487, USA ¶ Chemical

Science Division and Center for Nanophase Materials Sciences, Oak Ridge National

Laboratory, Oak Ridge, TN 37831, USA

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 34

ABSTRACT: Propylene epoxidation with molecular oxygen has been proposed as a green and alternative process to produce propylene oxide (PO). In order to develop catalysts with high selectivity, high conversion and long stability for the direct propylene epoxidation with molecular oxygen, understandings of catalyst structure and reactivity relationships are needed. Here, we combined atomic layer deposition and deposition precipitation to synthesize series of well-defined Au-based catalysts to study the catalyst structure and reactivity relationships for propylene epoxidation at 373 K. We showed that by decorating TiO2 on gold surface the inverse TiO2/Au/SiO2 catalysts maintained ~90% selectivity to PO regardless the weight loading of the TiO2. The inverse TiO2/Au/SiO2 catalysts exhibited improved regeneration compared to Au/TiO2/SiO2. The inverse TiO2/Au/SiO2 catalysts can be regenerated in 10% oxygen at 373 K, while the Au/TiO2/SiO2 catalysts failed to regenerate at as high as 473 K. Combined characterizations of the Au-based catalysts by X-ray absorption spectroscopy, scanning transmission electron microscopy and UV-vis spectroscopy suggested that the unique selectivity and regeneration of TiO2/Au/SiO2 derives from the site-isolated Ti sites on Au surface and AuSiO2 interfaces which are critical to achieve high PO selectivity and generate only coke-like species with high oxygen content. The high oxygen content coke-like species can therefore be easily removed. These results indicate that inverse TiO2/Au/SiO2 catalyst represents a system capable of realizing sustainable gas phase propylene epoxidation with molecular oxygen at low temperature.

ACS Paragon Plus Environment

2

Page 3 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

1. INTRODUCTION Propylene oxide (C3H6O, PO) is a chemical intermediate of great value for the production of many useful materials, including polyol, propylene glycol, and glycol ethers.

1

Current

industrial PO production such as chlorohydrin and hydroperoxide processes generates chlorinated or peroxycarboxylic byproducts which pose environmental issues. Direct epoxidation of propylene to propylene oxide with oxygen (O2) represents an alternative, clean, and potentially more efficient route.2-7 Haruta and co-workers first discovered that nm-sized gold clusters on TiO2 are active and highly selective for the direct epoxidation of propylene with O2.2, 8-9

It was proposed that when molecular propylene, hydrogen and oxygen were co-fed, peroxide

species could be formed on the surface of gold, and subsequently oxidize propylene that adsorbed on TiO2 sites to propylene oxide. It was later realized that small gold nanoparticles (< 5 nm) and isolated Ti active sites are necessary for achieving high activity and selectivity towards propylene oxide. Propylene molecules that adsorb on adjacent Ti sites could lead to formation of unwanted byproducts such as CO2. For this reason, gold nanoparticles supported by titanium silicalite (Au/TS-1) are the most studied catalysts for the propylene epoxidation reaction as Ti sites are highly isolated in TS-1.7,

10-13

Despite studies over the past decades, the propylene

epoxidation reaction still comes short in propylene conversion, catalyst stability and hydrogen efficiency, and therefore, there is a need to build precise structure-performance relationship for Au-based catalysts for propylene epoxidation reaction and design new catalysts based on this relationship. In precious metal catalysts, metal surfaces are generally considered as the active sites where catalytic reactions take place. A loss of catalytic activity sometimes results from the strong metal support interaction (SMSI) in which the metal surface active sites gradually become

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 34

covered by the metal oxide support. However, it has been found that a thin oxide film grown on a metal can exhibit much higher catalytic activity than the metal substrate under the same reaction conditions,14-15 where a 5 Å FeO(111) film grown on Pt(111) is active for CO oxidation by O 2 at 450 K, a temperature far below that at which Pt(111) itself is active. In a combined experimental and theoretical study, inverse CeO2/Au(111) catalyst is also found to be highly efficient for water gas shift reaction.16 Thin metal oxide films often exhibit dramatically different chemical and physical properties compared to their bulk counterparts. In particular, if the oxide film is only a few angstroms thick and supported by a metal substrate, the underlying metal substrate often further affects the oxide film properties through charge transfer.17-20 The concept of the “electronic theory of catalysts” for tuning catalytic activity by the thickness and structure of oxide over metal surface was first developed in the 1950s.21 Despite the above studies, due to a lack of precise synthesis methods to control the thickness and coverage of the ultrathin (a few nanometers) oxide film overcoat for practical catalyst, the inverse catalyst studies are mostly carried out in surface science studies. Atomic layer deposition (ALD) is a thin film coating technique commercially used in the microelectronic industry.22-23 Per the international technology roadmap for semiconductors, the sub-10 nanometer sized transistor is expected be in commercial mass production by 2018. This trend is expected to be further enhanced by 5 nm sized transistor technology by 2020. It seems that the two disparate industries, namely semiconductor manufacturing and heterogeneous catalysis, have gradually converged to the same length scale. The high-k/metal gate materials deposited by ALD play an important role in minimizing the ever smaller dimension of the modern process nodes. ALD’s capability to tailor nanomaterials at the atomic scale makes it a promising, alternative method to design and prepare well-defined nano catalysts.24-27 ALD has

ACS Paragon Plus Environment

4

Page 5 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

since been utilized to synthesized nanostructured catalysts such as single atom catalysts,28-29 inverse oxide/metal catalysts,30-33 bimetallic catalysts,34-36 and functionalized zeolites37-39. In this work, we carefully alter the TiO2 ALD cycles and their sequence on Au-based catalysts for the propylene epoxidation reaction. The resulting TiO2 decorated catalysts provides a well-defined system for studying the effects of TiO2 geometry and loading on the stability and regeneration ability of the Au-based catalysts, as well as the activity and selectivity of the propylene epoxidation at temperatures as low as 373 K. 2. EXPERIMENTAL 2.1 Catalyst synthesis The composition and structure of the catalysts were carefully designed and tuned by adjusting the sequence of Au and TiO2 deposited on silica gel. We categorize these catalysts as Au/ncTiO2/SiO2 or inverse ncTiO2/Au/SiO2 when TiO2 or gold was first deposited, respectively. The italic n represents the number of TiO2 ALD cycles. 2.1.1 Synthesis of Au/ncTiO2/SiO2 Catalysts The Silicycle S10040M SiO2 with a surface area of ~100 m2/g was used as the support. TiO2 ALD was performed in a viscous flow benchtop reactor (Gemstar-6, Arradiance). Ultrahigh purity N2 (Airgas, 99.999%) was used as carrier gas and further purified using a Supelco gas purifier (Sigma-Aldrich) before entering the reactor. 500 mg of SiO2 was uniformly spread onto a stainless steel tray with a mesh cover on top of it. The mesh can prevent the spill of the samples but still provides efficient diffusion of the ALD precursors and product gases in and out of the tray. The time sequence for one cycle ALD can be expressed as t1-t2-t3-t4, where t1 and t3

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 34

correspond to the exposure times of the two precursors, and t2 and t4 are the nitrogen purge times between two precursors. The TiO2 was deposited on high surface area SiO2 gel at a temperature of 473 K using alternating exposure to titanium isopropoxide (TTIP, Sigma-Aldrich, 99.999%) and deionized water, with a time sequence of 25s-200s-3.75s-200s.40 The TTIP was stored in a sealed stainless steel bottle at 353 K to provide sufficient vapor pressure. Four samples were prepared using 1, 3, 5 and 10 cycles of TiO2 ALD over SiO2. The deposition of gold on TiO2 ALD modified SiO2 supports was carried out by means of deposition precipitation (DP).7 Approximately 0.2 g of HAuCl4·3H2O (Sigma Aldrich, 99.99%) was added to 100 ml deionized water, followed by the addition of 1 g of support. NaOH (1mol/L) was drop-wise added to the solution to keep the pH at ~7 for 2 h. The solid portion was separated from the solution and washed with 50 ml of deionized water, separated again, and finally dried in air at room temperature overnight. 2.1.2. Synthesis of ncTiO2/Au/SiO2 Catalysts The precursor Au(en)2Cl3 was synthesized by slowly adding 0.9 mL of ethylenediamine (en, Sigma-Aldrich, 99.5%) to an aqueous solution of HAuCl4·3H2O (2.0 g in 20.0 mL of H2O) until a transparent brown solution was formed.41 The transparent brown solution was stirred for 30 min. Subsequently, 140.0 mL of ethanol was added. A precipitation was immediately produced. The solution was stirred for another 40 min. The final product was filtered, washed by 160 mL ethanol, and dried overnight in oven at 313 K. 0.27 g of Au(en)2Cl3 was dissolved in 150 mL of H2O. The pH value of the solution was adjusted to 10.0 by the addition of the NaOH (1 mol/L) solution. Subsequently, 5.25 g of the silica gel was added. The pH value of the solution decreased immediately, and then the pH value

ACS Paragon Plus Environment

6

Page 7 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

was adjusted to 10.0 by addition of the NaOH solution. The mixed solution was stirred for 2 h. The solid was separated from the solution and washed by 50 ml deionized water, separate again and finally dried in vacuum oven for 5h. TiO2 ALD was performed in the same way mentioned before. Four samples were prepared using 1, 3, 5 and 10 TiO2 ALD cycles over Au/SiO2. 2.2 Characterization In situ Quartz Crystal Microbalance (QCM) measurements were carried out under ALD conditions in real time to establish the TiO2 ALD recipe and to understand the surface chemistry. The QCM system contains a Maxtek BSH-150 bakeable sensor, AT-cut quartz sensor crystals (Colorado Crystal Corporation), and a Maxtek TM400 film thickness monitor. Once the TiO 2 ALD recipe was established, various ALD cycles of TiO2 thin films were deposited on Si(100) wafers. The thickness of the TiO2 thin films was measured using a J.A.Woollam alpha-SE variable angle spectroscopic ellipsometer (VASE). The data were obtained over a wavelength of 380-900 nm and an incidence angle of 70o. The time sequence for operating one ALD cycle of TiO2 on planar quartz sensor crystals and Si(100) wafers is 1s-5s-1s-5s. Scanning electron microscopy (SEM) measurements were carried out using a bench-top microscope (Hitachi TM-1000). The high angle annular dark field scanning transmission electron microscopy (HAADF-STEM) imaging was performed using a FEI Tecnai F-20 transmission electron microscope equipped with a 200 keV thermal Schottky Field emission gun. The size histogram of gold nanoparticles was determined by measuring the diameter of more than 250 particles for each catalyst. X-ray diffraction (XRD) measurement was carried out using Rigaku MiniFlex 600 powder X-ray diffractometer using Cu K-alpha radiation operated at 40 kV and 15

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 34

mA; 2 theta data from 2ºto 90ºwere obtained at a scanning speed of 0.5 o/min. The BrunauerEmmett-Teller (BET) surface area was measured via nitrogen adsorption at 77 K by Micromeritics Gemini 275 system. The UV-Vis diffuse reflectance spectra (DRS) of the fresh, spent and regenerated samples were obtained using a Cary 5000 UV/Vis spectrophotometer equipped with a praying mantis kit at room temperature. The amount of gold deposited on the support was determined by inductively coupled plasma mass spectrometry (ICP-MS). Fourier transform infrared (FTIR) spectroscopy measurements were performed using PerkinElmer FT-IR Spectrum Two spectrophotometer with diamond attenuated total reflectance (ATR) attachment. With the average of 8 repetitious scans, the scanning was conducted from 4000 to 400 cm-1. Prior to measurement, the samples were placed evenly and pressed to contact with the ATR crystal to obtain adequate band intensities. From the pressure readout, the same force was applied to acquire quantitative analysis for SiO2 concentration. X-ray absorption spectroscopy (XAS) measurements, including extended X-ray absorption fine structure spectroscopy (EXAFS) and X-ray absorption near edge structure spectroscopy (XANES), were conducted at the Au L3 edge (~11919 eV) at beamline 10-BM of the Materials Research Collaborative Access Team (MRCAT) at the Advanced Photon Source (APS) at Argonne National Laboratory. The XAS spectra were recorded in the transmission mode. The samples were pressed into a cylindrical holder which can hold six samples simultaneously. The loading of the sample was optimized to achieve a step height of 1. In a typical measurement, the catalyst samples were first fully reduced in 50 ml/min 3.5% H2 in helium at 523 K for 30 min and subsequently cooled to room temperature in ultra-high purity helium. EXAFS regime data fittings were performed using WINXAS 3.1. A single shell model fit of the EXAFS data was obtained between k = 3.0−12.0 Å−1 and r = 1.6−3.5 Å, respectively.

ACS Paragon Plus Environment

8

Page 9 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

2.3 Measurement of Catalytic Activity The catalytic reaction for gas-phase propylene epoxidation was carried out in a quartz tubular reactor with a diameter of 10 mm at normal pressure. The quartz tube reactor was placed within a furnace using approximately 0.15 g catalyst of 60-80 mesh size. Approximately 1.5 g quartz sand of 25-35 mesh size was used to dilute the catalyst to improve the temperature uniformity. A blank experiment was performed under the same reaction conditions using only the quartz sand. A thermocouple was attached on the quartz tube wall to directly measure the reaction temperature. The reactant mixture consisted of 10/10/10/70 vol% of hydrogen (Airgas, 99.999%), oxygen (Airgas, 99.999%), propylene (Matheson, 99.9%), and argon (Airgas, 99.999%), with a total flow of 35 ml/min equivalent to a gas hourly space velocity (GHSV) of 14,000 ml h-1 g-1cat. Without any pretreatment of the catalyst, the epoxidation reaction was carried out at 373 ± 2 K, which was reached by a temperature ramp rate of 1.5 K/min. The effluent line was wrapped with heating tape and heated to above 80 °C to prevent condensation of all the byproducts. The experimental conditions for kinetic studies are listed in Table S1. To extract the oxygen order, the oxygen concentration was changed solely at 453 K. The changes in propylene and hydrogen concentrations were used to extract propylene and hydrogen orders. Since the reaction order is independent with temperature, same reaction order for each reactant was applied to derive the reaction rate constant k at 413 K, 433 K, and 453 K. When hydrogen peroxide was used as a reactant, a syringe pump equipped with a borosilicate syringe through a Teflon tube was used to directly introduce H2O2 (30 wt%, Sigma Aldrich, containing stabilizer) into the reactor. To minimize the H2O2 decomposition, the syringe

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 34

and Teflon tubing were wrapped with aluminum foil and their length minimized. The partial pressure of H2O2 could be adjusted by changing the pumping speed. A bed of quartz wool around 4 cm was placed above the catalyst bed where H2O2 fully evaporated. Based on the assumption that H2O2 can be fully decomposed to H2O and O2 through the heated stainless steel reactor, a blank run with only H2O2 in argon was performed to measure the weight percentage of the H2O2 solution. The H2O2 weight percentage calculated from O2 in the effluent is around 33±2%, which is similar to the specification. The concentration of the reactants and products was measured by an online gas chromatograph (GC, Agilent 5890) equipped with a flame ionization detector (FID) and a thermal conductivity detector (TCD). Oxygen, hydrogen, and propylene were separated using Carboxen-1010 PLOT capillary column (length 30 m; i.d. 0.53 mm) and analyzed via TCD, while propylene oxide, ethanal, acetone, propanal and acrolein were separated using a RT-UBOND column (length 30 m; i.d. 0.53 mm) and analyzed via FID. Carbon monoxide (CO) and carbon dioxide (CO2) were not detected under the reaction conditions used in this study. The propylene conversion and selectivity were calculated directly as follows: Conversion = moles of (C3-oxygenates + 2/3ethanal)/moles of propylene in the feed … eqn (1) PO selectivity = moles of PO/moles of (C3-oxygenates + 2/3ethanal+CO2/3) … eqn (2) 3. RESULTS AND DISCUSSIONS 3.1 TiO2 Atomic Layer Deposition For the purpose of precisely controlling the loading of TiO2 using ALD, the growth rate of TiO2 ALD was carefully studied using an in situ quartz crystal microbalance (QCM) housed inside the ALD chamber, and was later confirmed using ex situ spectroscopic ellipsometry

ACS Paragon Plus Environment

10

Page 11 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

measurements. The in situ QCM results in Figure 1a show a mass gain of 0.06 μg/cm2 per TiO2 ALD cycle, equivalent to a growth rate of 0.3 Å per TiO2 ALD cycle when TTIP and H2O were used as the reactants at 473 K. The linear growth rate indicates that the ALD growth indeed follows a surface self-limiting nature. An expanded view of the mass gain is illustrated in Figure 1b. The surface received +0.08 μg/cm2 (chemisorption of –Ti(OCH(CH3)2)2 ligands) and –0.04 μg/cm2 mass gain –OCH(CH3)2 ligands replaced by hydroxyl groups) during TTIP and H2O exposure, respectively, which is consistent with the literature.40, 42 A series of TiO2 thin films, using 50, 100, 200 and 300 cycles of TiO2 ALD, respectively, were prepared on clean Si(100) wafers. The thickness of TiO2 increases linearly as a function of TiO2 ALD cycles with a growth rate of 0.3 Å as determined by ex situ spectroscopic ellipsometry in Figure 1c. Both QCM and ellipsometry results are consistent with each other.

Figure 1. In situ QCM studies of TiO2 ALD at 473 K for (a) the first 100 cycles and (b) magnified growth rate (blue line – TTIP pulse; red line – H2O pulse). (c) Growth rate of TiO2

ACS Paragon Plus Environment

11

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 34

ALD at 473 K determined by ex situ spectroscopic ellipsometry. Due to its unique feature of self-limiting surface reactions, the growth rate obtained from the wafers was used to estimate the TiO2 surface coverage on the SiO2 gel support using equations (3) and (4). The SEM image of the bare SiO2 gel is shown in Figure S1. The coverage θ is obtained from the equations below, 𝜃=

𝑑 ⋯ 𝑒𝑞𝑛 (3) đ

3 𝑀 đ = √ ⋯ 𝑒𝑞𝑛 (4) 𝜌𝑁

where d is the deposited TiO2 thickness, đ is TiO2 monolayer thickness in nm, M is the molar mass of TiO2 (79.87 g/mol), ρ is the density of TiO2 (4.23 × 10-21 g/nm3), and N is the number of atoms per mole (6.02 × 1023/mol). In this work, catalysts were coated by 1, 3, 5, and 10 TiO2 ALD cycles, respectively. Their estimated thickness, coverage, and weight loading are shown in Table 1.

If we assume that one ALD cycle of TiO2 are site-isolated and homogeneously

dispersed on SiO2 surface, their intermolecular distance is estimated to be ~13.6 Å. BET surface area measurements were carried out before and after 10 cycles of TiO2 ALD. The N2 adsorptiondesorption isotherms exhibit typical H4-type hysteresis, indicating the presence of mesopores. The ALD coating did not alter the surface area or pore volume or pore size of the SiO 2 support (please see Figure S2). Table 1. Estimated thickness, coverage and weight loading after TiO2 ALD treatment. (The monolayer thickness đ is equal to 3.2 Å) Cycle number of TiO2 ALD 1

Estimated thickness (Å) 0.3

Coverage θ (%) 9.7

Estimated TiO2 loading (wt%) 1.3

ACS Paragon Plus Environment

12

Page 13 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

3 5 10 Figure 2 is the FTIR spectra

0.9 1.6 3.1 of SiO2 modified

29.1 48.4 96.9 by different ALD

3.8 6.4 12.7 cycles of TiO2. The

spectra exhibited three dominant features, and these are assigned as: the low frequency band centered at 457 cm-1 formed by the rocking motions of the oxygen atoms perpendicular to the SiO-Si plane; the intermediate frequency band centered at 800 cm-1 is due to the symmetrical stretching of the oxygen atoms along the line bisecting the axis formed by the Si-O-Si; the high frequency band centered at 1072 cm-1 is attributed to the antisymmetric stretching of the oxygen atoms along the direction parallel to Si-Si.43 The intensity of those dominant peaks decreased as TiO2 ALD cycles increased, which corresponded to the increased of TiO2 loading. In addition, the peak position of the Si-O-Si antisymmetric stretching vibration shifted to higher wavenumber (blueshift, from 1072 to 1082 cm-1) with the increase of TiO2 incorporation, which implies that the Si-O-Si bond angle increased and/or the Si-O bond length decreased.44

Figure 2. ATR-FTIR transmittance spectra of SiO2 (green), 1cTiO2/SiO2 (red), 3cTiO2/SiO2 (blue), 5cTiO2/SiO2 (pink), and 10cTiO2/SiO2 (orange). 3.2 Characterizations

ACS Paragon Plus Environment

13

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 34

Figure 3 shows the STEM images of the as-prepared samples where homogeneously dispersed gold nanoparticles on the titanium coated SiO2 support. Unfortunately, it is unlikely that STEM can provide useful information on TiO2, since the catalyst support SiO2 gel is a high surface area porous material, which is considered as not well-defined for atomic resolution STEM. In addition, the low loading and low z-number of Ti make it more difficult to be distinguished from the SiO2 background. The average diameter of the as-prepared gold nanoparticles is 3.6 ± 0.6 nm and 4.1 ± 0.5 nm for the as-prepared Au/3cTiO2/SiO2 and Au/10cTiO2/SiO2 catalysts, respectively. Au nanoparticles appear as bright dots on the TiO2 coated SiO2 support. ICP-MS measurements showed that the amount of gold were ~0.05 wt% and ~0.7 wt% for Au/ncTiO2/SiO2 series and ncTiO2/Au/SiO2 series, respectively.

(a)

(b)

(c)

(d)

ACS Paragon Plus Environment

14

Page 15 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 3. STEM images of the as-prepared Au/3cTiO2/SiO2 (a, b) and Au/10cTiO2/SiO2 (c, d) catalysts. Figure 4 shows the XRD patterns for the as-prepared 10cTiO2/SiO2, the as-prepared Au/10cTiO2/SiO2, and the Au/10cTiO2/SiO2 calcined at 600 °C for 2h. The broad peaks at 2θ range from 20ºto 30º, and they are representative peaks for an amorphous SiO2 support. Neither TiO2 nor Au features were clearly observed in the XRD patterns of the as-prepared Au/10cTiO2/SiO2. The missing features seem to suggest that both TiO2 and Au are highly dispersed on the SiO2 support, which is desired for highly active propylene epoxidation catalysts. The Au/10cTiO2/SiO2 catalyst was calcined in air at 600 °C for 2 h to artificially increase the gold particle diameter. Indeed, the calcined Au/10cTiO2/SiO2 catalysts show sharp diffraction peaks that are consistent with the face-centered cubic (fcc) gold structure.

ACS Paragon Plus Environment

15

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 34

Figure 4. XRD patterns for 10cTiO2/SiO2 (black), Au/10cTiO2/SiO2 (red), and Au/10cTiO2/SiO2 calcined at 600 °C for 2h (blue).

To understand the impact of TiO2 loading and location on the morphological and electronic structure of the Au catalysts, Au L3 XAS measurements were carried out. The Au catalysts were in a metallic oxidation state (please see XANES spectra in Figure S3). Regardless the location of TiO2, Fourier transform of EXAFS spectra of Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 are very similar, as shown in Figure 5 (a). Both spectra showed peaks in the magnitude of the Fourier transform for the first shell Au-Au at about 2.3 Å and 3.0 Å and a shoulder peak at about 2.0 Å. Detailed EXAFS first shell fittings are concluded in Table 2 and the fitting quality is shown in Figure S4 and S5. The Au-Au bond distance of the gold nanoparticles is about 2.81 Å – 2.85 Å, slightly contracted from the Au-Au bond distance of 2.88 Å in a bulk Au. A contracted bond distance is widely observed in precious nanoparticles and is caused by the contraction of the under-coordinated surface atoms to increase the coordination numbers of the surface atoms. 45-46 The Au-Au coordination numbers are well below their bulk value of 12 which is expected for a bulk fcc structure, indicating that the gold is in the form of small nanoparticles. Diameter of the Au nanoparticles can be estimated using their coordination numbers using an empirical equation suggested by Miller and co-workers.47 The results are listed in Table 2. The Au-Au coordination and interatomic Au-Au bond length exhibits a linear relationship within the range of our study, as shown in Fig 5 (b). It seems to suggest that the structure of gold is independent of the location of TiO2. The diameters of gold nanoparticles of all the samples are smaller than the STEM results. We ascribed the difference to the low resolution of the conventional STEM

ACS Paragon Plus Environment

16

Page 17 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

that we used. The estimated diameter of Au in the ncTiO2/Au/SiO2 catalysts is slightly larger than the Au/ncTiO2/TiO2 catalysts. The difference in diameter is probably caused by slightly different methods used to prepare Au nanoparticles on TiO2/SiO2 and SiO2 surface, respectively. Due to the mismatch between the isoelectric point of SiO2 (IEP=2) and pH used to hydrolyze the Au precursors (HAuCl4), Au(en)2Cl3 had to be used to prepare Au nanoparticles on the SiO2 surface.

Figure 5. (a) Fourier transform of EXAFS. (b) Correlation between the CNAu-Au and Au-Au bond distance for a series of catalysts Au/ncTiO2/SiO2 (black square) and ncTiO2/Au/SiO2 (red dot).

Table 2. EXAFS fit parameters for Au-Au scatterers. (k2: Δk = 3-12 Å-1 and Δr = 1.6-3.5 Å) Samples

CNAu-Au

R (Å)

DWF (x10-3)

E0 (eV)

Est. Dia. (nm)

5.9

2.83

0.001

-1.7

1.3

Au/ncTiO2/SiO2 Au/1cTiO2/SiO2

ACS Paragon Plus Environment

17

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 34

Au/3cTiO2/SiO2

6.3

2.83

0.001

-1.9

1.6

Au/5cTiO2/SiO2

5.6

2.81

0.001

-2.3

1.1

Au/10cTiO2/SiO2

7.1

2.83

0.001

-3.0

2.1

1cTiO2/Au/SiO2

8.1

2.84

0.001

-1.7

2.7

3cTiO2/Au/SiO2

7.4

2.84

0.001

-1.7

2.3

5cTiO2/Au/SiO2

8.3

2.85

0.001

-1.5

2.9

ncTiO2/Au/SiO2

CN is coordination numbers with error ±10%. R is bond distance with error ±0.02 Å. DWF is Debye-Waller factor. E0 is energy shift.

3.3 Catalyst Performance in Propylene Epoxidation 3.3.1 PO Selectivity Table S2 lists the selectivity for all the byproducts. There is no CO or CO2 detected in this study as their concentration is likely below the detection limit of the TCD detector. The PO selectivity decreased with increasing TiO2 ALD cycles for the Au/ncTiO2/SiO2 catalysts as shown in Figure 6. The selectivity to PO decreased from 95% to 63% when the TiO2 ALD cycle increased from 1 to 10 cycles. In the previous propylene epoxidation studies, isolated Ti sites are believed to be critical to obtain high selectivity for propylene epoxidation reaction as the adjacent Ti sites can generate byproducts such as acrolein, propanal and acetone.48 Our results are consistent with this observation in the literature. According to previous studies on TiO2 ALD, TTIP precursors react with the surface hydroxyl groups and generate isolated Ti sites in the first cycle. In the subsequent TiO2 ALD cycles, the TTIP precursors will react with both the hydroxyl groups on open Si sites to generate more isolated Ti sites, as well as the hydroxyl groups on the

ACS Paragon Plus Environment

18

Page 19 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Ti sites to generate adjacent Ti sites. The overall effect is that the isolated sites decrease with increasing TiO2 ALD cycles, leading to decreasing PO selectivity. Interestingly, the PO selectivity of the ncTiO2/Au/SiO2 seems to be less affected by TiO2 ALD cycles as shown in Figure 6(b). The selectivity to PO decreases slightly from ~91% to ~88% when the TiO2 ALD cycle increases from 1 to 10 cycles, which seems to suggest that the TiO2 deposited on Au remains mostly isolated. It is useful to notice that the TiO2 ALD growth rate also depends on the density of surface active sites, e.g., surface hydroxyl groups in this case. The real growth rate of TiO2 ALD on Au could be significantly lower than TiO2 ALD on SiO2 due to much lower density of surface OH on Au. The low coverage of OH on Au could be helpful to keep the TiO2 site isolated. Surface science studies of the synthesis of TiO2 nanoparticles on Au(111) surface showed that TiO2 forms islands at a coverage as low as 0.05 ML,49 however, Ti was generated using physical vapor deposition (PVD) on a clean surface without hydroxyl groups. The deposited titanium metal first formed Au-Ti alloys on the surface of Au. TiO2 was subsequently formed by exposing the system to O2. In the case of TiO2 ALD, it is likely that TTIP reacts with the surface hydroxyl groups near Au or at Au/SiO2 interface forming Au-O-Ti bonds which can enhance the thermal stability of the system and help Ti sites to remain isolated. Such increased thermal stability is similar to surface science studies of model Au catalysts on hydroxylated surface.50 It is known that TiO2 can be formed on Au nanoparticles’ surface even with a very small number of TiO2 ALD cycles. For example, it was reported that the CO chemisorption dramatically decreased after depositing five ALD cycles of TiO2 on Au/Al2O3 surface.51

ACS Paragon Plus Environment

19

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 34

Figure 6. PO selectivity for series of catalysts (a) Au/ncTiO2/SiO2 and (b) ncTiO2/Au/SiO2. n=1, (black line), n=3 (red line), n=5 (blue line), and n=10 (pink line). (c) Average PO selectivity for series of catalysts Au/ncTiO2/SiO2 (black line) and ncTiO2/Au/SiO2 (red line) under different cycle of TiO2 ALD modification. 3.3.2 PO formation rate In terms for PO formation rate, five ALD cycles of TiO2 seems to offer the highest PO formation rate in both Au/ncTiO2/SiO2 and ncTiO2/Au/SiO2 series, representing the optimal synergic effects between isolated Ti sites and Au. For Au/ncTiO2/SiO2 catalysts (Figure 7a), the PO formation rate increases with increasing number of TiO2 ALD cycles from 1 to 5. We ascribed the increasing PO formation rate to the increasing number of Au and Ti pairs. Accompanied with a significant increase in byproduct selectivity (please see Table S2), the PO formation rate dramatically decreased when 10 ALD cycles of TiO2 were used, due to the loss of isolated Ti sites. For ncTiO2/Au/SiO2 series (Figure 7b), TiO2 gradually covers the surface of gold nanoparticles when the TiO2 ALD cycles increases from 1 to 10, leading to an increase of Ti sites and a decrease of Au sites. 3 – 5 ALD cycles of TiO2 on Au nanoparticles seems to

ACS Paragon Plus Environment

20

Page 21 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

provide an optimized balance between Ti and Au sites. Note that Au/SiO2 was also studied as a controlled experiment. No reactivity was observed.

Figure 7. PO formation rate for series of catalysts (a) Au/ncTiO2/SiO2 and (b) ncTiO2/Au/SiO2. n=1, (black line), n=3 (red line), n=5 (blue line), and n=10 (pink line). 3.3.3 Deactivation and regeneration Unfortunately, the Au-based catalysts studied in this work all suffered from a certain degree of deactivation when they were evaluated at 373 K. The time-on-stream stability is shown in Figure 8. After being tested for 10 h under reaction conditions, Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 catalysts deactivated by 48.6% and 54.8%, respectively. In an attempt to regenerate both catalysts, the spent catalysts were regenerated in 10% O2 at 373 K for 2 h. The treatment successfully regenerated the overcoated 1cTiO2/Au/SiO2 to its initial performance but failed to regenerate Au/1cTiO2/SiO2. An additional attempt to regenerate Au/1cTiO2/SiO2 catalyst at a higher temperature (473 K) was also unsuccessful. To confirm the above

ACS Paragon Plus Environment

21

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 34

observations, repeated experiments were performed to test and regenerate both Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 catalysts. Similar phenomena were observed.

Figure 8. PO formation rate for catalyst (a) Au/1cTiO2/SiO2, (b) 1cTiO2/Au/SiO2, and (c) 1cTiO2/SiO2. There are three possible reasons for gold-based catalysts deactivation in propylene epoxidation, namely sintering of gold nanoparticles, carbon deposition (coking) on the active sites and the rearrangement of active TiO2 sites. Figure 9 shows the STEM images of the spent catalysts Au/3cTiO2/SiO2 and Au/10cTiO2/SiO2. The average size of observable gold nanoparticles was 2.9 and 4.0 nm, respectively, similar to their diameter before the catalyst evaluation. XRD patterns were also measured on the spent catalysts immediately after the catalyst evaluation. The XRD patterns of the spent catalysts (data not shown) are similar to those of the as-prepared catalysts where no peaks ascribed to fcc Au was observed, indicating small gold nanoparticles. In a control experiment, hydrogen peroxide (H2O2) and propylene (1:1 molar

ACS Paragon Plus Environment

22

Page 23 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

ratio) were co-fed to evaluate the 1cTiO2/SiO2 catalyst. The catalyst 1cTiO2/SiO2 deactivated by 52.9%. Treatment in 10% O2 at 373 K for 2 h did not regenerate the catalyst, suggesting that the deactivation is probably due to factors other than sintering of gold nanoparticles.

(a)

(b)

Figure 9. STEM images of spent catalysts (a) Au/3cTiO2/SiO2 and (b) Au/10cTiO2/SiO2. The insets show the gold nanoparticle size distributions. Another possible reason for catalyst deactivation is the TiO2 rearrangement during reaction. This rearrangement can be reflected by the change of the Ti coordination numbers which was investigated using UV-vis DRS spectroscopy. The absorption spectra were recorded for fresh, deactivated, and regenerated catalysts as shown in Figure 10 for Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2. The spectra in each of those three catalysts showed the same trend which excludes the change in TiO2 and gold geometry as one of the reasons for catalysts deactivation and regeneration. The strong absorptions at 200 to 220 nm are assigned to isolated tetrahedral Ti in the framework.52 The band around 230 and 275 nm are attributed to the presence of Ti in fivefold and six-fold coordination,53 respectively, while that at ~305 nm corresponds to anatase

ACS Paragon Plus Environment

23

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 34

TiO2.54 The spectra show an absorption in the visible range between 500 to 600 nm assigned to the localized surface plasmon resonance of Au nanoparticles.55

Figure 10. UV-vis DR spectra for samples (a) Au/1cTiO2/SiO2 and (b) 1cTiO2/Au/SiO2 As a result, it seems that the possible reason for the deactivation is the residual carbon blocking the catalyst active sites. Coke-like species were previous observed using in situ Raman spectroscopy56 and IR spectroscopy57-58, respectively. These species have been assigned as strongly adsorbed oligomerization or polymerization of propylene or propylene oxide,59 bidentate propoxy species,57 and carbonates/carboxylates48. The true identity of the coke-like species is yet to be determined. One possible explanation is that the coke-like organic species generated by the 1cTiO2/Au/SiO2 contain a relatively high oxygen content and are easy to be burned off, as compared to the possible low-oxygen content coke species produced by Au/1cTiO2/SiO2.

3.3.4 Kinetic Studies

ACS Paragon Plus Environment

24

Page 25 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Kinetic studies of Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 catalysts were performed by varying the concentration of the reactants and reaction temperatures. The detailed experimental conditions listed in Table S1 follow those used by Taylor et al.60 Twenty three sets of experiments were carried out to determine the reaction order and rate constant for each catalysts. The experimental results are presented in Table S3 and S4. Unlike the experiments operated at low temperature (373 K), little deactivation was observed during the kinetic measurements performed above 413 K, suggesting that the carbon-species were removed by oxygen under reaction conditions. From the previous studies, PO has strong interaction with the catalyst surface and considered to have negative impact on catalyst performance due to active sites blocking.61 Therefore, PO concentration would be expected in powder rate law. Because of the relatively low propylene conversion, PO concentration is approximately two orders of magnitude less than the concentration of each reactant. Moreover, no obvious catalyst deactivation observed, which indicates that interaction between PO and catalyst surface could be ignored. Hence, it is reasonable to neglect the formed PO concentration in the power rate law. Those two catalysts were fit individually using the linear regression features. To keep a linear model, ln(k) rather than k was fitted. The power rate law converted to below equation, ln(rPO) = ln(k)+αln[H2]+βln[O2]+γln[C3H6] … eqn (5) The overall data fitting is shown in Figure S6 where all the data points overlap or close to the parity lines. Figure S7 shows the percent residual randomly distributed around the zero lines within ±10%. The derived fitting parameters were shown in Table 3. For catalyst Au/1cTiO2/SiO2, PO formation rate was found to be most dependent on hydrogen concentration, whereas oxygen has the highest effect on PO rate for catalyst 1cTiO2/Au/SiO2. The rate law of propylene epoxidation using Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 catalysts are similar to what

ACS Paragon Plus Environment

25

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 34

Delgass and coworkers observed on Au/TS-1 catalysts where they obtained a rate law rPO = k[H2]0.60[O2]0.31[C3H6]0.18.60 The apparent activation energy is derived based on the Arrhenius plot shown in Figure 11. The activation energy Ea for Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 catalysts is 24.1 kJ/mol and 38.7 kJ/mol, respectively, comparable to the literature value (25~36 kJ/mol).62 These results seem to suggest that the active sites in ALD modified Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 catalysts are similar to those in Au/TS-1.

Table 3. Power rate law parameters obtained Catalyst

α

β

γ

k(140)a

k(160)

k(180)

Ea b

Au/1cTiO2/SiO2

0.52

0.33

0.22

14.22

22.55

30.48

24.1±1.9

1cTiO2/Au/SiO2

0.39

0.49

0.26

25.35

46.05

74.71

38.7±2.0

rPO = k(T)[H2]α[O2]β[C3H6]γ, where k(T)=A𝑒 (−𝐸𝑎/𝑅𝑇) a

Unit: gPO/h•kgcat•atmα+β+γ

b

Unit: kJ/mol

ACS Paragon Plus Environment

26

Page 27 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 11. Arrhenius plot for propylene epoxidation using Au/1cTiO2/SiO2 and 1cTiO2/Au/SiO2 catalysts, respectively. Reaction condition: 10% O2, 10% H2, and 10% C3H6 in Ar

4. Conclusions In conclusion, we used ALD to carefully synthesized Au/TiO2/SiO2 and inverse TiO2/Au/SiO2 catalysts used in propylene epoxidation with molecular oxygen to propylene oxide. Our study showed that inverse TiO2/Au/SiO2 enabled high PO selectivity (~90%) and easy regeneration at 373 K, results not found in Au/TiO2/SiO2. The high PO selectivity independent of TiO2 ALD cycles indicated that the ALD TiO2 remained as site-isolated on Au surface and Au-SiO2 interfaces. Furthermore, combined STEM and UV-Vis spectroscopy studies suggested that the deactivation of the gold-based catalysts were due to deposition of coke-like species. The repeated, easy regeneration of the inverse TiO2/Au/SiO2 at low temperature is proposed to result from high oxygen content coke-like species that can be removed at low temperature (373 K). It is interesting to note that similar deactivation and failure to regenerate were observed on TiO2/SiO2 catalyst used in propylene epoxidation with hydrogen peroxide at

ACS Paragon Plus Environment

27

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 34

373 K. It is envisioned that a potentially energy efficient, low temperature propylene epoxidation can be achieved by using the inverse TiO2/Au/SiO2 catalyst. Kinetic measurements at temperature above 413 showed little catalysts deactivation. The reaction rate equations determined in our work are similar to propylene epoxidation using Au/TS-1 catalysts.

ASSOCIATED CONTENT Supporting Information. Additional details on SEM images, BET measurements, XANES spectra, EXAFS data fitting, and catalytic performance. This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Author *Tel.:1-256-824-6527. E-mail: [email protected]. Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This work is sponsored by the National Science Foundation (Grant # CBET-1511820 and CBET1510485). The authors thank Mr. Johnny Goodwin of University of Alabama for taking the STEM images. Part of the work including the BET surface area measurements and UV-Vis spectroscopy were conducted at the Center for Nanophase Materials Sciences, which is a DOE Office of Science User Facility. MRCAT operations are supported by the Department of Energy and the MRCAT member institutions. This research used resources of the Advanced Photon

ACS Paragon Plus Environment

28

Page 29 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DE-AC0206CH11357.

REFERENCES (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)

Weissermel, K.; Arpe, H.-J., Industrial Organic Chemistry. Fourth ed.; WILEY-VCH: Weinheim, Germany, 2003. Hayashi, T.; Tanaka, K.; Haruta, M. Selective Vapor-Phase Epoxidation of Propylene over Au/TiO2 Catalysts in the Presence of Oxygen and Hydrogen. J. Catal. 1998, 178, 566-575. Lei, Y.; Mehmood, F.; Lee, S.; Greeley, J.; Lee, B.; Seifert, S.; Winans, R. E.; Elam, J. W.; Meyer, R. J.; Redfern, P. C.; et al. Increased Silver Activity for Direct Propylene Epoxidation via Subnanometer Size Effects. Science 2010, 328, 224-228. Marimuthu, A.; Zhang, J.; Linic, S. Tuning Selectivity in Propylene Epoxidation by Plasmon Mediated Photo-Switching of Cu Oxidation State. Science 2013, 339, 15901593. Khatib, S. J.; Oyama, S. T. Direct Oxidation of Propylene to Propylene Oxide with Molecular Oxygen: A Review. Catal. Rev.-Sci. Eng. 2015, 57, 1-39. Nijhuis, T. A.; Makkee, M.; Moulijn, J. A.; Weckhuysen, B. M. The Production of Propene Oxide: Catalytic Processes and Recent Developments. Ind. Eng. Chem. Res. 2006, 45, 3447-3459. Lee, W.-S.; Akatay, C.; Stach, E. A.; Ribeiro, F. H.; Delgass, W. N. Gas-Phase Epoxidation of Propylene in the Presence of H2 and O2 over Small Gold Ensembles in Uncalcined TS-1. J. Catal. 2014, 313, 104-112. Uphade, B. S.; Yamada, Y.; Akita, T.; Nakamura, T.; Haruta, M. Synthesis and Characterization of Ti-MCM-41 and Vapor-Phase Epoxidation of Propylene using H2 and O2 over Au/Ti-MCM-41. Appl. Catal., A 2001, 215, 137-148. Uphade, B. S.; Akita, T.; Nakamura, T.; Haruta, M. Vapor-Phase Epoxidation of Propene using H2 and O2 over Au/Ti-MCM-48. J. Catal. 2002, 209, 331-340. Gaudet, J.; Bando, K. K.; Song, Z.; Fujitani, T.; Zhang, W.; Su, D. S.; Oyama, S. T. Effect of Gold Oxidation State on the Epoxidation and Hydrogenation of Propylene on Au/TS-1. J. Catal. 2011, 280, 40-49. Huang, J. H.; Lima, E.; Akita, T.; Guzman, A.; Qi, C. X.; Takei, T.; Haruta, M. Propene Epoxidation with O2 and H2: Identification of the Most Active Gold Clusters. J. Catal. 2011, 278, 8-15. Ferrandez, D. M. P.; Fernandez, I. H.; Teley, M. P. G.; de Croon, M.; Schouten, J. C.; Nijhuis, T. A. Kinetic Study of the Selective Oxidation of Propene with O2 over Au-Ti Catalysts in the Presence of Water. J. Catal. 2015, 330, 396-405. Feng, X.; Sheng, N.; Liu, Y. B.; Chen, X. B.; Chen, D.; Yang, C. H.; Zhou, X. G. Simultaneously Enhanced Stability and Selectivity for Propene Epoxidation with H 2 and

ACS Paragon Plus Environment

29

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29)

Page 30 of 34

O2 on Au Catalysts Supported on Nano-Crystalline Mesoporous TS-1. ACS Catal. 2017, 7, 2668-2675. Sun, Y. N.; Qin, Z. H.; Lewandowski, M.; Carrasco, E.; Sterrer, M.; Shaikhutdinov, S.; Freund, H. J. Monolayer Iron Oxide Film on Platinum Promotes Low Temperature CO Oxidation. J. Catal. 2009, 266, 359-368. Fu, Q.; Li, W. X.; Yao, Y. X.; Liu, H. Y.; Su, H. Y.; Ma, D.; Gu, X. K.; Chen, L. M.; Wang, Z.; Zhang, H.; et al. Interface-Confined Ferrous Centers for Catalytic Oxidation. Science 2010, 328, 1141-1144. Rodriguez, J. A.; Ma, S.; Liu, P.; Hrbek, J.; Evans, J.; Perez, M. Activity of CeOx and TiOx Nanoparticles Grown on Au(111) in the Water-Gas Shift Reaction. Science 2007, 318, 1757-1760. Giordano, L.; Pacchioni, G.; Goniakowski, J.; Nilius, N.; Rienks, E. D. L.; Freund, H. J. Charging of Metal Adatoms on Ultrathin Oxide Films: Au and Pd on FeO/Pt(111). Phys. Rev. Lett. 2008, 101, 026102. Kulawik, M.; Nilius, N.; Rust, H. P.; Freund, H. J. Atomic Structure of Antiphase Domain Boundaries of a Thin Al2O3 Film on NiAl(110). Phys. Rev. Lett. 2003, 91, 256101. Sterrer, M.; Risse, T.; Pozzoni, U. M.; Giordano, L.; Heyde, M.; Rust, H. P.; Pacchioni, G.; Freund, H. J. Control of the Charge State of Metal Atoms on Thin MgO Films. Phys. Rev. Lett. 2007, 98, 096107. Senanayake, S. D.; Stacchiola, D.; Rodriguez, J. A. Unique Properties of Ceria Nanoparticles Supported on Metal: Novel Inverse Ceria/Copper Catalysts for CO Oxidation and the Water-Gas Shift Reaction. Acc. Chem. Res. 2013, 46, 1702-1711. Cabrera, N.; Mott, N. F. Theory of the Oxidation of Metals. Reports on Progress in Physics 1949, 12, 163-184. George, S. M. Atomic Layer Deposition: An Overview. Chem. Rev. 2010, 110, 111-131. Zaera, F. The Surface Chemistry of Atomic Layer Depositions of Solid Thin Films. J. Phys. Chem. Lett. 2012, 3, 1301-1309. Singh, J. A.; Yang, N.; Bent, S. F. Nanoengineering Heterogeneous Catalysts by Atomic Layer Deposition. Annu. Rev. Chem. Biomol. Eng. 2017, 8, 41-62. Lu, J.; Elam, J. W.; Stair, P. C. Atomic Layer Deposition-Sequential Self-Limiting Surface Reactions for Advanced Catalyst "Bottom-Up" Synthesis. Surf. Sci. Rep. 2016, 71, 410-472. Lu, M.; Nuwayhid, R. B.; Wu, T.; Lei, Y.; Amine, K.; Lu, J. Atomic Layer Deposition for Lithium-Based Batteries. Advanced Materials Interfaces 2016, 3, 1600564. O'Neill, B. J.; Jackson, D. H. K.; Lee, J.; Canlas, C.; Stair, P. C.; Marshall, C. L.; Elam, J. W.; Kuech, T. F.; Dumesic, J. A.; Huber, G. W. Catalyst Design with Atomic Layer Deposition. ACS Catal. 2015, 5, 1804-1825. Piernavieja-Hermida, M.; Lu, Z.; White, A.; Low, K.-B.; Wu, T.; Elam, J. W.; Wu, Z.; Lei, Y. Towards ALD Thin Film Stabilized Single-Atom Pd1 Catalysts. Nanoscale 2016, 8, 15348-15356. Sun, S.; Zhang, G.; Gauquelin, N.; Chen, N.; Zhou, J.; Yang, S.; Chen, W.; Meng, X.; Geng, D.; Banis, M. N.; et al. Single-Atom Catalysis Using Pt/Graphene Achieved through Atomic Layer Deposition. Scientific Reports 2013, 3, 1775.

ACS Paragon Plus Environment

30

Page 31 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(30) (31) (32) (33) (34)

(35) (36) (37)

(38)

(39) (40) (41) (42) (43)

Lei, Y.; Lee, S.; Low, K.-B.; Marshall, C. L.; Elam, J. W. Combining Electronic and Geometric Effects of ZnO-Promoted Pt Nanocatalysts for Aqueous Phase Reforming of 1-Propanol. ACS Catal. 2016, 6, 3457-3460. Lu, J. L.; Fu, B. S.; Kung, M. C.; Xiao, G. M.; Elam, J. W.; Kung, H. H.; Stair, P. C. Coking- and Sintering-Resistant Palladium Catalysts Achieved through Atomic Layer Deposition. Science 2012, 335, 1205-1208. Onn, T. M.; Zhang, S. Y.; Arroyo-Ramirez, L.; Chung, Y. C.; Graham, G. W.; Pan, X. Q.; Gorte, R. J. Improved Thermal Stability and Methane-Oxidation Activity of Pd/Al2O3 Catalysts by Atomic Layer Deposition of ZrO2. ACS Catal. 2015, 5, 5696-5701. Gould, T. D.; Lubers, A. M.; Corpuz, A. R.; Weimer, A. W.; Falconer, J. L.; Medlin, J. W. Controlling Nanoscale Properties of Supported Platinum Catalysts through Atomic Layer Deposition. ACS Catal. 2015, 5, 1344-1352. Lei, Y.; Liu, B.; Lu, J. L.; Lobo-Lapidus, R. J.; Wu, T. P.; Feng, H.; Xia, X. X.; Mane, A. U.; Libera, J. A.; Greeley, J. P.; et al. Synthesis of Pt-Pd Core-Shell Nanostructures by Atomic Layer Deposition: Application in Propane Oxidative Dehydrogenation to Propylene. Chem. Mater. 2012, 24, 3525-3533. Wang, X. F.; Zhao, H. Y.; Wu, T. P.; Liu, Y. Z.; Liang, X. H. Synthesis of Highly Dispersed and Highly Stable Supported Au-Pt Bimetallic Catalysts by a Two-Step Method. Catal. Lett. 2016, 146, 2606-2613. Cao, K.; Liu, X.; Zhu, Q. Q.; Shan, B.; Chen, R. Atomically Controllable Pd@Pt CoreShell Nanoparticles towards Preferential Oxidation of CO in Hydrogen Reactions Modulated by Platinum Shell Thickness. Chemcatchem 2016, 8, 326-330. Wu, Y.; Lu, Z.; Emdadi, L.; Oh, S. C.; Wang, J.; Lei, Y.; Chen, H.; Tran, D. T.; Lee, I. C.; Liu, D. Tuning External Surface of Unit-Cell Thick Pillared MFI and MWW Zeolites by Atomic Layer Deposition and its Consequences on Acid-Catalyzed Reactions. J. Catal. 2016, 337, 177-187. Xu, D.; Wu, B. S.; Ren, P. J.; Wang, S. Y.; Huo, C. F.; Zhang, B.; Guo, W. P.; Huang, L. H.; Wen, X. D.; Qin, Y.; et al. Controllable Deposition of Pt Nanoparticles into a KL Zeolite by Atomic Layer Deposition for Highly Efficient Reforming of n-heptane to Aromatics. Catal. Sci. Tech. 2017, 7, 1342-1350. Gong, T.; Qin, L. J.; Lu, J.; Feng, H. ZnO Modified ZSM-5 and Y Zeolites Fabricated by Atomic Layer Deposition for Propane Conversion. Phys. Chem. Chem. Phys. 2016, 18, 601-614. Ritala, M.; Leskela, M.; Niinisto, L.; Haussalo, P. Titanium Isopropoxide as a Precursor in Atomic Layer Epitaxy of Titanium Dioxide Thin Films. Chem. Mater. 1993, 5, 11741181. Wu, Z. L.; Zhou, S. H.; Zhu, H. G.; Dai, S.; Overbury, S. H. DRIFTS-QMS Study of Room Temperature CO Oxidation on Au/SiO2 Catalyst: Nature and Role of Different Au Species. J. Phys. Chem. C 2009, 113, 3726-3734. Lei, Y.; Liu, B.; Lu, J.; Libera, J. A.; Greeley, J. P.; Elam, J. W. Effects of Chlorine in Titanium Oxide on Palladium Atomic Layer Deposition. J. Phys. Chem. C 2014, 118, 22611-22619. Almeida, R. M.; Pantano, C. G. Structural Investigation of Silica-Gel Films by InfraredSpectroscopy. J. Appl. Phys. 1990, 68, 4225-4232.

ACS Paragon Plus Environment

31

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(44) (45) (46) (47) (48) (49) (50) (51)

(52) (53)

(54) (55)

(56) (57) (58)

Page 32 of 34

Kim, Y. H.; Hwang, M. S.; Kim, H. J.; Kim, J. Y.; Lee, Y. Infrared Spectroscopy Study of Low-Dielectric-Constant Fluorine-Incorporated and Carbon-Incorporated Silicon Oxide Films. J. Appl. Phys. 2001, 90, 3367-3370. Lei, Y.; Zhao, H. Y.; Diaz-Rivas, R.; Lee, S.; Liu, B.; Lu, J. L.; Stach, E. A.; Winans, R. E.; Chapman, K. W.; Greeley, J. P.; et al. Adsorbate-Induced Structural Changes in 1-3 nm Platinum Nanoparticles. J. Am. Chem. Soc. 2014, 136, 9320-9326. Lei, Y.; Jelic, J.; Nitsche, L. C.; Meyer, R.; Miller, J. T. Effect of Particle Size and Adsorbates on the L3, L2 and L1 X-ray Absorption Near Edge Structure of Supported Pt Nanoparticles. Top. Catal. 2011, 54, 334-348. Miller, J. T.; Kropf, A. J.; Zha, Y.; Regalbuto, J. R.; Delannoy, L.; Louis, C.; Bus, E.; van Bokhoven, J. A. The Effect of Gold Particle Size on Au-Au Bond Length and Reactivity toward Oxygen in Supported Catalysts. J. Catal. 2006, 240, 222-234. Nijhuis, T. A.; Sacaliuc-Parvulescu, E.; Govender, N. S.; Schouten, J. C.; Weckhuysen, B. M. The Role of Support Oxygen in the Epoxidation of Propene over Gold-Titania Catalysts Investigated by Isotopic Transient Kinetics. J. Catal. 2009, 265, 161-169. Biener, J.; Farfan-Arribas, E.; Biener, M.; Friend, C. M.; Madix, R. J. Synthesis of TiO2 Nanoparticles on the Au(111) Surface. J. Chem. Phys. 2005, 123, 094705. Brown, M. A.; Fujimori, Y.; Ringleb, F.; Shao, X.; Stavale, F.; Nilius, N.; Sterrer, M.; Freund, H. J. Oxidation of Au by Surface OH: Nucleation and Electronic Structure of Gold on Hydroxylated MgO(001). J. Am. Chem. Soc. 2011, 133, 10668-10676. Wang, C.; Wang, H.; Yao, Q.; Yan, H.; Li, J.; Lu, J. Precisely Applying TiO2 Overcoat on Supported Au Catalysts Using Atomic Layer Deposition for Understanding the Reaction Mechanism and Improved Activity in CO Oxidation. J. Phys. Chem. C 2016, 120, 478-486. Sinha, A. K.; Seelan, S.; Akita, T.; Tsubota, S.; Haruta, M. Vapor Phase Propylene Epoxidation over Au/Ti-MCM-41 Catalysts Prepared by Different Ti Incorporation Modes. Appl. Catal., A 2003, 240, 243-252. Sinha, A. K.; Seelan, S.; Okumura, M.; Akita, T.; Tsubota, S.; Haruta, M. ThreeDimensional Mesoporous Titanosilicates Prepared by Modified Sol-Gel Method: Ideal Gold Catalyst Supports for Enhanced Propene Epoxidation. J. Phys. Chem. B 2005, 109, 3956-3965. Yap, N.; Andres, R. P.; Delgass, W. N. Reactivity and Stability of Au in and on TS-1 for Epoxidation of Propylene with H2 and O2. J. Catal. 2004, 226, 156-170. Tsukamoto, D.; Shiraishi, Y.; Sugano, Y.; Ichikawa, S.; Tanaka, S.; Hirai, T. Gold Nanoparticles Located at the Interface of Anatase/Rutile TiO 2 Particles as Active Plasmonic Photocatalysts for Aerobic Oxidation. J. Am. Chem. Soc. 2012, 134, 63096315. Nijhuis, T. A.; Visser, T.; Weckhuysen, B. M. Mechanistic Study into the Direct Epoxidation of Propene over Gold/Titania Catalysts. J. Phys. Chem. B 2005, 109, 1930919319. Mul, G.; Zwijnenburg, A.; van der Linden, B.; Makkee, M.; Moulijn, J. A. Stability and Selectivity of Au/TiO2 and Au/TiO2/SiO2 Catalysts in Propene poxidation: An in situ FTIR study. J. Catal. 2001, 201, 128-137. Bravo-Suarez, J. J.; Bando, K. K.; Lu, J.; Haruta, M.; Fujitani, T.; Oyama, S. T. Transient Technique for Identification of True Reaction Intermediates: Hydroperoxide Species in

ACS Paragon Plus Environment

32

Page 33 of 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(59) (60) (61) (62)

Propylene Epoxidation on Gold/Titanosilicate Catalysts by X-ray Absorption Fine Structure Spectroscopy. J. Phys. Chem. C 2008, 112, 1115-1123. Nijhuis, T. A.; Huizinga, B. J.; Makkee, M.; Moulijn, J. A. Direct Epoxidation of Propene Using Gold Dispersed on TS-1 and Other Titanium-Containing Supports. Ind. Eng. Chem. Res. 1999, 38, 884-891. Taylor, B.; Lauterbach, J.; Blau, G. E.; Delgass, W. N. Reaction Kinetic Analysis of the Gas-Phase Epoxidation of Propylene over Au/TS-1. J. Catal. 2006, 242, 142-152. Sinha, A.; Seelan, S.; Tsubota, S.; Haruta, M. Catalysis by Gold Nanoparticles: Epoxidation of Propene. Top. Catal. 2004, 29, 95-102. Taylor, B.; Lauterbach, J.; Delgass, W. N. Gas-Phase Epoxidation of Propylene over Small Gold Ensembles on TS-1. Appl. Catal., A 2005, 291, 188-198.

ACS Paragon Plus Environment

33

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 34

TOC Graphic

ACS Paragon Plus Environment

34